Historically, the study of the effects of environmental variables on the corrosion behavior of metals has been performed by varying a single species at a time. At most, a two or three variable matrix of tests has been performed. In this study, a statistically designed matrix of tests was performed to examine the main-term effects and several two-factor interaction terms of 15 environmental variables present in ground-water solutions on the corrosion behavior of carbon steel. Cyclic potentiodynamic polarization experiments were performed to determine several characteristic polarization parameters. Regression analyses established which of the environmental variables had a statistically significant effect on each of the polarization parameters. The environmental variables were selected to simulate basalt rock ground-water solutions with the emphasis being the application of carbon steel as a high-level nuclear waste container (overpack). It was found that, to accurately describe the effects of ground-water variables, two-factor interaction terms were often as important, or more so, than the main-effect terms. A limitation of the technique is that, with the large concentration ranges used, the design concentrations could not be achieved due to ion interaction; this produced a correlation among factors in the regression model and inaccuracies in the predictions.